Paper Reading Angelic Debugging



Speed bug the repair candidate search by value based reasoning.

If replacing an expression e with a value v fixes the failing test, the repairing e has the potential to fix the failing test. If replacing e with a value v different than e computes does not break break amassing test, then we have some freedom to change e. We say that it is flexible and a good repair candidate.


Used to debug several programs with known defects.


I found it hard to understand what big idea I can learn. How is value-based reasoning a new thing, an inspiring thing?

Practical Value

What you can learn from this to make your research better?

The assumptions that authors made makes this paper feasible, yet compelling in terms of motivation.

Details and Problems From the presenters’ point of view, what questions might audience ask?

Important assumption/constraint: We consider only those faulty programs in which the defat can be repaired by altering at most one expression.

Another important assumption: Programmer will first locate a scope to localize a fault.

Important note: this system is not generating the patch; However, the replacement value it generates might inspire the programmer to choose the right patch.